Fig. 5: Performance of our deep learning approach versus other models. | Communications Engineering

Fig. 5: Performance of our deep learning approach versus other models.

From: A deep learning framework based on structured space model for detecting small objects in complex underwater environments

Fig. 5

Two different test sets were selected to validate the performance and generalization ability of various models in different scenarios. a AP50-95 values of the models on the Test A dataset. b AP50 values of the models on the Test A dataset. c AP50-95 values of the models on the Test B dataset. d AP50 values of the models on the Test B dataset.

Back to article page